ravi-garg / unsupervised_depth_estimation Goto Github PK

I got the latest version caffe and set WITH_PYTHON_LAYER := 1 in makefile.config. However, I still get the following error:

===

F0325 16:49:50.756420 6840 layer_factory.hpp:81] Check failed: registry.count(type) == 1 (0 vs. 1) Unknown layer type: Warping (known types: AbsVal, Accuracy, ArgMax, BNLL, BatchNorm, BatchReindex, Bias, Clip, Concat, ContrastiveLoss, Convolution, Crop, Data, Deconvolution, Dropout, DummyData, ELU, Eltwise, Embed, EuclideanLoss, Exp, Filter, Flatten, HDF5Data, HDF5Output, HingeLoss, Im2col, ImageData, InfogainLoss, InnerProduct, Input, LRN, LSTM, LSTMUnit, Log, MVN, MemoryData, MultinomialLogisticLoss, PReLU, Parameter, Pooling, Power, Python, RNN, ReLU, Reduction, Reshape, SPP, Scale, Sigmoid, SigmoidCrossEntropyLoss, Silence, Slice, Softmax, SoftmaxWithLoss, Split, Swish, TanH, Threshold, Tile, WindowData)

===
Is there anything I missed?

What should the argument to net.forward() be in pycaffe.

can this train model be run in python?

Hello, I'm trying to load a traiend network and use it to depth estimation, however the file train_iter_40000.caffemodel requires a .prototxt' to be imported in matlab using importCaffeNetwork. I tried with the .prototxt files such as resnet50.prototxt, but it says Check if your network contains layers which are not defined in caffe.proto..
Thanjs for your help.

Hi,

I downloaded your prediction result mat, but I can not recover it correctly. I got a 142x697 image from predictions in mat. I plot it our but it's quite weird. All the images look like this example:

Is there anything I have to do before I save the data to png?

Hi,Thanks for your great sharing!
I want to generate the depth of kitti ,I have no idea about it even if i have read the development kit provided by the website(http://www.cvlibs.net/datasets/kitti/raw_data.php). could you please share your code about generating the depth iamges?
Thank you！

i see the 'KITTI_vis_test_seq.m' that you offered in the program, you seem to set (focal f * distance B) equals to 389.6304，i know the B is 0.54 meter in the kitti dataset paper, but how/where do you get the camera`s focal f?

hi,
how soon can you make the code available?